Behavioral pharmacology of the pyrazoloquinoline CGS 9896, a novel putative anxiolytic

CGS 9896, a pyrazoloquinoline that displaces [³H]benzodiazepines from their CNS receptors, has been claimed to be a nonsedative anxiolytic in animals. However, conflicting results have been obtained with this compound in animal tests of anxiety and seizures. The present study reports the effects of CGS 9896 in three animal tests of anxiety, against seizures induced by picrotoxin and pentylenetetrazole, and on the holeboard test in rats. CGS 9896 (injected intraperitoneally) has an anxiolytic-like activity in the social interaction test (5--10 mg/kg) and in the elevated plus-maze (10--20 mg/kg) but not in the Vogel punished drinking procedure (5--10 mg/kg). CGS 9896 had potent anticonvulsant effects against pentylenetetrazole (total protection at 5 mg/kg) but was considerably less potent against picrotoxin-induced seizures (only 50% reduction at 40 mg/kg). CGS 9896 (10 mg/kg) caused significant reductions in spontaneous exploratory and locomotor behavior in the holeboard test in rats. In all test procedures, CGS 9896 was considerably less potent than the benzodiazepines diazepam or chlordiazepoxide. In conclusion, these data support the evidence suggesting that CGS 9896 possesses anxiolytic and anticonvulsant properties in rodents but suggest that caution should be observed in concluding that a compound is nonsedative on the basis of one or two measures.     

CGS 9896: A nonbenzodiazepine,nonsedating potential anxiolytic

Various behavioral and neurochemical studies indicate that CGS 9896 may represent a novel, nonsedating anxiolytic. This substance, chemically related to the benzodiazepine antagonist CGS 8216, was effective in conflict and nonconflict models of anxiety. At the same time, CGS 9896 did not disrupt rotorod performance or decrease levels of responding in various operant procedures. In fact, CGS 9896 reversed the deficit in rotorod behavior produced by diazepam. CGS 9896 did not generalize to diazepam in rats trained to discriminate diazepam from vehicle. However, rats trained to discriminate CGS 9896 from vehicle generalized classical benzodiazepines to CGS 9896. These results suggest an anxioselective effect associated with CGS 9896 discriminative stimuli. Preliminary studies suggest that this pyrazoloquinoline does not produce dependence. Neurochemical analysis reveals that CGS 9896 binds avidly to benzodiazepine receptors both in vitro and in vivo. However, the binding characteristics of this compound differ from classical benzodiazepines in various respects. Two alternative hypotheses are discussed that may explain the behavioral and neurochemical differences between CGS 9896 and classical benzodiazepines.     

Lack of tolerance or withdrawal effects in mice after chronic administration of the non-sedating anxiolytic, CGS 9896

CGS 9896, a non-sedating anxiolytic, was compared to diazepam with respect to the development of tolerance and withdrawal. Both compounds were administered daily to mice at various doses (3, 10 or 30 mg/kg) for periods of up to 4 weeks. Measures of sedation/muscle relaxation, motor activity and anticonvulsant effects were then assessed. When administered acutely, CGS 9896 increased motor activity, had no effect on traction reflex, and elevated the threshold for PTZ-induced convulsions. After chronic administration of CGS 9896, no changes in these parameters were observed compared to the effects seen after acute treatment. Acute administration of diazepam reduced motor activity, impaired traction reflex and increased PTZ-induced convulsion threshold. Tolerance developed to the effects of diazepam in all three measures. Following a four week dosing period with 30 mg/kg of either CGS 9896 or diazepam, the drugs were withdrawn and similar behavioral measures obtained at various withdrawal intervals up to 15 days. In separate groups of mice, precipitated withdrawal was also assessed by the administration of the benzodiazepine agonist, CGS 8216. No effects were observed after any period of withdrawal from CGS 9896. By contrast, withdrawal from diazepam resulted in significant alterations of motor activity and convulsion threshold. These results indicate that CGS 9896 is likely to be free of undesirable tolerance and withdrawal effects typically associated with the benzodiazepines.     

The pentylenetetrazol-like interoceptive stimulus produced by ethanol withdrawal is potentiated by bicuculline and picrotoxinin

We investigated whether the interoceptive discriminative stimulus (IDS) arising from ethanol withdrawal was related to decreased activity of the gamma-aminobutyric acid (GABA) system by determining whether the sensitivity of rats to the GABA antagonists was altered by chronic treatment with ethanol. Rats were trained to obtain food reward by responding on one lever following pentylenetetrazol (PTZ) and the other lever following saline. Whereas all of the trained rats selected the PTZ-appropriate lever after PTZ, no more than 50% of them selected this lever following an optimum dose of bicuculline or picrotoxinin. After either saline or diazepam (5 mg/kg), all of the rats selected the saline-appropriate lever. Ethanol (0.24 mol/kg/day) was then administered to the rats for 4 days via a nutritionally balanced liquid diet. Between 48 and 96 hours postethanol, 30% of the rats selected the PTZ-appropriate lever following saline, whereas selection of this lever was increased to 80% following either bicuculline or picrotoxinin. Thus, further antagonism of GABAergic activity increased the subjective effect of ethanol withdrawal. These data support the hypothesis that the PTZ-like IDS produced during withdrawal from ethanol is related to an ethanol-induced deficit in the activity of the GABA-benzodiazepine receptor-coupled chloride channel.     

CGS 8216, a benzodiazepine antagonist, reduces food intake in food-deprived rats

CGS 8216, a benzodiazepine receptor antagonist with weak inverse agonist properties, reduced food intake in food-deprived rats when administered orally or intraperitoneally at doses that antagonize diazepam. This effect was sustained when CGS 8216 was administered daily for five days, indicating no rapid tolerance to the anorectic effect. Ro 15-1788 did not reduce feeding when administered orally, and was active only at high intraperitoneal doses (54 and 100 mg/kg). CGS 9896, a close analog of CGS 8216 but a benzodiazepine partial agonist with anxiolytic properties, did not reduce food intake at doses as high as 100 mg/kg IP or PO. These results support prior suggestions that benzodiazepine receptors may modulate feeding behavior, and suggest that CGS 8216 may have appetite suppressant properties.     

Inosine and N6-substituted adenosine analogs lack anxiolytic activity in the pentylenetetrazol discrimination model of anxiety

Although some in vitro studies have raised the possibility that endogenous purines mediate the therapeutic effects of the benzodiazepines, no behavioral studies have been performed to confirm or reject this hypothesis. Consequently, inosine and the adenosine A1 receptor agonists N6- (L-phenylisopropyl) adenosine (L-PIA) and 2-chloroadenosine (2CA) were evaluated for diazepam-like anxiolytic activity in the pentylenetrazol-saline discrimination model of anxiety. Rats were trained to press one of two levers for food reward after pentylenetetrazol (PTZ) injection (20 mg/kg) and to press the other after saline injection. During testing in these rats, diazepam (10 mg/kg) blocked the PTZ-induced selection of the PTZ-appropriate lever (i.e., rats selected the saline-correct lever). Inosine, L-PIA, and 2CA neither blocked selection of the PTZ-appropriate lever after PTZ administration nor did they reverse the diazepam blockade. It is suggested that adenosine or inosine may not act endogenously to mediate the anxiolytic effects of benzodiazepines.     

Anxiolytic action of CGS 9896 on mouse exploratory behavior

A new non-benzodiazepine compound proposed as a non-sedating anxiolytic was tested in the mouse exploratory model of anxiety. CGS 9896 significantly increased the number of light dark transitions at doses beginning at 7.5 mg/kg i.p. Analysis of general locomotor activity at these doses revealed no change in spontaneous motor activity in a photocell equipped activity monitor. Pretreatment with the benzodiazepine receptor antagonist, Ro15-1788, 10 mg/kg i.p., blocked the increase in light dark transitions produced by CGS 9896. These data support the interpretation that CGS 9896 acts as an anxiolytic through the benzodiazepine receptor, and appears to have no sedating properties within the anxiolytic dose range.     

CGS 8216 and CGS 9896, novel pyrazoloquinoline benzodiazepine ligands with benzodiazepine agonist and antagonist properties

CGS 8216 and CGS 9896 are two recently described compounds which interact with benzodiazepine binding sites but have pharmacological, biochemical and behavioral characteristics which distinguish them from classical benzodiazepines. CGS 8216 shows properties of a weak inverse agonist, while CGS 9896 shows properties of a mixed agonist/antagonist. Experiments using quantitative autoradiography to determine benzodiazepine binding site interactions of these compounds in discrete anatomical areas are described. Results indicate that [3H]-CGS 8216 does not show any regional differentiation in binding characteristics in 7 brain areas studied. CGS 9896 preferentially inhibited [3H]-flunitrazepam from cerebellar sites compared to hippocampal dentate gyrus sites, but the magnitude of this effect was small. These data support the conclusion that CGS 9896 is acting preferentially at putative benzodiazepine type 1 sites and is consistent with the mixed agonist/antagonist profile of the compound.     

Distinctive behavioral effects of the pyrazoloquinoline CGS 8216 in squirrel monkeys

The behavioral effects of the pyrazoloquinoline CGS 8216 were studied in squirrel monkeys trained to respond under a fixed-interval (FI) schedule of food presentation. Dose-effect curves were determined by administering cumulative doses IV during timeout periods that preceded sequential components of the FI schedule. CGS 8216 (0.1-3.0 mg/kg) produced dose-related decreases in the rate of FI responding. In comparison, diazepam (0.1-3.0 mg/kg) had biphasic effects under identical conditions: intermediate doses increased the rate, whereas high doses decreased the rate of FI responding. Pretreatment with the benzodiazepine antagonist Ro 15-1788 (3.0 or 5.6 mg/kg) attenuated the decreases in response rate normally produced by high doses of CGS 8216. The behavioral effects of CGS 8216 were not altered systematically by pretreatment with either diazepam (0.3-3.0 mg/kg) or the alpha 2-adrenergic agonist clonidine (0.01-0.03 mg/kg). The results suggest that CGS 8216 has benzodiazepine inverse agonist effects on schedule-controlled behavior of squirrel monkeys. CGS 8216 can, however, be distinguished from inverse agonists of the beta-carboline type on the basis of its effects in the presence of diazepam or clonidine.     

Behavioral effects of CGS 8216 alone,and in combination with diazepam and pentobarbital in dogs

Beagle dogs (N=3) responded under a multiple fixed-interval (FI) 300 sec, fixed-ratio (FR) 30 schedule of food presentation. The pyrazoloquinoline derivative CGS 8216, given either intravenously (0.01–3.0 mg/kg) or orally (0.1–30.0 mg/kg) had little effect on either the rate or temporal pattern of responding during either component. Both diazepam (0.3 to 17.5 mg/kg, PO) and pentobarbital (0.1–17.5 mg/kg, PO) produced qualitatively similar effects on behavior. Rates of responding during the FI components first increased, then decreased with increasing doses; both drugs produced only dose-related decreases in the rate of responding during the FR components. CGS 8216 antagonized some of the behavioral effects of diazepam; FI and FR response rates returned to baseline, however the effects of diazepam on quarter-life values were not appreciably altered by CGS 8216. The effects of pentobarbital on schedule-controlled responding were not antagonized by CGS 8216. These results indicate CGS 8216 is a selective benzodiazepine antagonist that does not produce benzodiazepine-like behavioral effects.     

Quinolines and anxiety: anxiogenic effects of CGS 8216 and partial anxiolytic profile of PK 9084

The effects of three quinoline derivatives, PK 8165 07.5 mg/kg). PK 9084 (7.5 mg/kg) and CGS 8216 (1 and 10 mg/kg), alone and in combination with chlordiazepoxide (5 mg/kg), were investigated in the social interaction test of anxiety. PK 8165 had no effect on social interaction but both PK 8165 and PK 9084 in combination with chlordiazepoxide produced sedation. PK 9084 exhibited a partial anxiolytic profile reflected by a drug x light interaction in the familiar test conditions. CGS 8216 (1 mg/kg) showed no significant effects on social interaction, but did counteract the sedative effect of chlordiazepoxide. The 10 mg/kg dose of CGS 8216 reduced social interaction between pairs of animals in familiar test conditions which is indicative of an anxiogenic effect. These intrinsic anxiogenic properties of CGS 8216 demonstrate that it cannot be considered an inert benzodiazepine antagonist.     

Pentylenetetrazole-like stimulus is produced in rats during withdrawal from ingested chlordiazepoxide

The present study was undertaken to determine whether withdrawal from chlordiazepoxide administered via a liquid diet would produce a pentylenetetrazole (PTZ)-like stimulus. Rats were trained with food reward in a two-lever operant task. Presses on one lever were reinforced after injections of PTZ (20 mg/kg, i.p.) and on the other lever after saline (1 ml/kg, i.p.). After rats had acquired the PTZ discrimination, training was halted, and chlordiazepoxide (240 mg/kg/day) was administered via a nutritionally balanced liquid diet to three groups of rats for 3, 4, or 6 days. Upon termination of chronic administration, withdrawal was precipitated with the benzodiazepine receptor blocker flumazenil (Ro 15-1788) given intraperitoneally. During precipitated withdrawal, the rats selected the PTZ-appropriate lever, indicating the presence of a PTZ-like stimulus, and this stimulus was blocked by phenobarbital (80 mg/kg, i.p.). The percentage of rats selecting the PTZ-appropriate lever depended on the duration of chlordiazepoxide treatment and dose of flumazenil. At 10 days after the last chlordiazepoxide dose, the rats had recovered baseline discrimination, as indicated by their selection of the saline appropriate lever following saline injections and the PTZ-appropriate lever following PTZ. These data indicate that a subjective effect of withdrawal similar to that produced by the anxiogenic drug PTZ is present during withdrawal from oral chlordiazepoxide.     

Lorazepam and pentobarbital discrimination: interactions with CGS 8216 and caffeine

Baboons and rats were trained under a two-lever, food-reinforced drug discrimination procedure. The training drug was either lorazepam (1.0 mg/kg) or pentobarbital (5.6 mg/kg in baboons, 10.0 mg/kg in rats). Under test conditions, a range of training drug doses occasioned 100% drug lever responding. CGS 8216 (3.2-10.0 mg/kg) combined with lorazepam produced a complete shift to the no-drug lever in both species; this shift was surmountable with higher doses of lorazepam. CGS 8216 (32.0 mg/kg) combined with pentobarbital produced a statistically significant decrease in drug-lever responding in rats, and in baboons CGS 8216 initially, but not subsequently, produced a complete shift to the no-drug lever. Caffeine (0.32-10.0 mg/kg) combined with lorazepam inconsistently decreased drug-lever responding across multiple determinations in baboons and significantly decreased drug lever responding in rats. Caffeine combined with pentobarbital also yielded an inconsistent decrease in drug lever responding in baboons but there was no effect in rats. Thus the most reliable and complete antagonism across species was obtained with the CGS 8216/lorazepam combinations.     

In vivo determination of efficacy of pyrazoloquinolinones at the benzodiazepine receptor

The pyrazoloquinolinones CGS 9896, CGS 9895 and CGS 8216 potently displace the benzodiazepines from their CNS binding site in vitro but have been reported to display agonist, partial agonist and antagonist activity respectively in a number of in vivo tests in rats. We found CGS 9896 to have only weak antipentylenetetrazole activity in mice at doses which produced near maximal displacement of [3H]flunitrazepam in vivo and were not able to detect any antipentylenetetrazole activity of CGS 9895. However, CGS 9895 blocked the anticonvulsant effect of diazepam at doses closely related to those which inhibited the in vivo binding of [3H]flunitrazepam and CGS 9896 also showed some reversal of the antipentylenetetrazole action of diazepam. This is consistent with the notion that CGS 9896 is a partial agonist and CGS 9895 an antagonist at the benzodiazepine receptor.     

Differential interactions between chlordiazepoxide, pentobarbital and benzodiazepine antagonists Ro 15-1788 and CGS 8216 in a drug discrimination procedure

Rats were trained to discriminate either chlordiazepoxide (CDP, 4 mg/kg, IP, N = 8) or pentobarbital (PB, 10 mg/kg, IP, N = 8) from saline in a two-lever food-reinforced procedure. CDP and PB dose-dependently substituted for each other (greater than or equal to 90% drug lever responses); indicating that their discriminative stimulus properties were closely similar. However, discriminative stimulus control induced by CDP and PB differentially was affected by the proposed benzodiazepine (BDZ) antagonists Ro 15-1788 (0.08-20 mg/kg, IP) and CGS 8216 (2.5-20 mg/kg, IP) in each experimental group; suggesting that the discriminative stimulus properties of CDP and PB are mediated by different mechanisms of action. When administered alone, Ro 15-1788 (5 and 20 mg/kg), but not CGS 8216, induced CDP like discriminative effects, suggesting that Ro 15-1788 may have partial (BDZ like) agonist properties, not shown by CGS 8216. Additional evidence for a behavioral difference between Ro 15-1788 and CGS 8216 is suggested by differential effects of both compounds on response rate. The results may reflect differential interactions of the compounds with the BDZ receptor-GABA receptor-Cl- ionophore complex.     

Discriminative stimulus properties of CGS 9896: interactions within the GABA/benzodiazepine receptor complex

Male rats were trained to discriminate the stimulus effects of CGS 9896 (30.0 mg/kg) from its vehicle. Once trained, discriminative performance was observed to be dose-responsive in the 3.75-30.0 mg/kg range and analysis of the dose-response curve generated an ED50 of 6.44 mg/kg. Generalization testing with chlordiazepoxide and pentobarbital produced CGS 9896-appropriate responding, whereas administration of the GABA agonists SL 75 102 resulted in 75% (intermediate) generalization to the CGS 9896 discriminative stimulus. Although full antagonism of the CGS 9896 cue was obtained following administration of Ro15-1788 and pentylenetetrazole, the inverse agonist DMCM failed to provide complete antagonism. These results suggest that the discriminative properties of CGS 9896 are consistent with its activity as a benzodiazepine receptor agonist.     

Discriminative stimulus properties of pentylenetetrazol and bemegride: Some generalization and antagonism tests

In an operant procedure of lever pressing on a FR 10 schedule of food reinforcement, male hooded rats were trained to respond with a lever on one side of a food cup following a drug injection, and to respond with a lever on the alternate side following a 1ml/kg saline injection. All of 14 subjects learned to discriminate reliably between the effects of 20 mg/kg pentylene-tetrazol (PTZ) and saline. Seven of eight rats learned to discriminate between the effects of bemegride (5 mg/kg) and saline. None of 14 rats learned to discriminate between 5 mg/kg PTZ and saline. The bemegride discriminative stimulus generalized to PTZ (20 mg/kg) and was antagonized by chlordiazepoxide (10 mg/kg). Chlordiazepoxide, diazepam, flurazepam, clobazam, and meprobamate were all effective antagonists of PTZ in a dose-dependent manner. Bemegride and cocaine generalized to the PTZ discriminative stimulus in a dose-dependent manner, but d-amphetamine, methylphenidate, and nicotine did not. Since bemegride and PTZ are convulsants at higher doses, the discriminative stimulus properties of these drugs might be based on a subtle convulsive brain state. The anxiolytic properties of benzodiazepines and meprobamate suggest that the discriminative stimulus produced by these convulsants is related to an anxiety-inducing action.     

Benzodiazepine-induced hyperphagia: Stereospecificity and antagonism by pyrazoloquinolines,CGS 9895 and CGS 9896

Non-deprived male rats were familiarised with a highly palatable diet until baseline consumption in a 30-min daily access period had stabilised. Stereospecificity of the hyperphagic effect of benzodiazepine receptor agonists was demonstrated using two enantiomers, the (S)-enantiomer being Ro11-3128 (methylclonazepam) and the (R)-enantiomer, Ro11-3624. The benzodiazepine receptor antagonists, Ro15-1788 and CGS 8216, reversed the hyperphagic effect of Ro11-3128. These data confirm the mediation of the hyperphagic effect of benzodiazepines by specific receptors. In further experiments, the effects of the pyrazoloquinolines CGS 9895 and CGS 9896 were examined both alone and also in combination with clonazepam. In doses of 1.25-10.0 mg/kg, neither CGS 9895 nor CGS 9896, when given alone, had a significant effect on the consumption of the palatable diet. Both, however, dose-dependently antagonised the hyperphagic effect of clonazepam. In a test of palatable food consumption, therefore, both compounds can be characterised as benzodiazepine antagonists.     

Discriminative stimulus properties of chlordiazepoxide and zolpidem. Agonist and antagonist effects of CGS 9896 and ZK 91296

In previous studies the effects of CGS 9896, a pyrazoloquinoline ligand at benzodiazepine receptors, in rats trained to discriminate benzodiazepines from vehicle, have been variable. The present experiment confirmed that this compound produced responding on the drug-lever in rats trained to discriminate 5 mg/kg of chlordiazepoxide from saline, and showed that CGS 9896 did not antagonise the effect of chlordiazepoxide in this test. In contrast, CGS 9896 antagonised the stimulus properties of zolpidem (2 mg/kg), a non-benzodiazepine hypnotic, which displaces benzodiazepines from their binding sites. The drug CGS 9896 also antagonised responding on the drug-lever produced by chlordiazepoxide in rats trained with zolpidem. The beta-carboline, ZK 91296, produced effects similar to those of CGS 9896, giving rise to responding on the drug-lever in rats trained with chlordiazepoxide and antagonising the zolpidem cue. These results demonstrate the mixed agonist-antagonist effects of CGS 9896 and ZK 91296 and suggest that the stimulus properties of chlordiazepoxide and zolpidem may be mediated by different sub-types of benzodiazepine receptors.     

Baclofen does not block interoceptive discriminative stimulus produced by pentylenetetrazol

Baclofen was compared with diazepam for ability to block the interoceptive discriminative stimulus produced by the anxiogenic drug pentylenetetrazol (PTZ). Male hooded rats of the Long Evans strain were trained in two-lever operant chambers to respond for food reward. Presses on one lever were reinforced after intraperitoneal injections of PTZ (20 mg/kg), and on the other lever after saline. After rats had acquired the PTZ discrimination, it was shown that pretreatment with baclofen failed to block the interoceptive stimulus produced by PTZ, whereas pretreatment with diazepam was effective. As baclofen is ineffective in a conflict procedure as well as in antagonizing the PTZ stimulus, but produces a diazepamlike interoceptive discriminative stimulus, it is concluded that elicitation of the diazepam stimulus may not be related to diazepam's anxiolytic efficacy and that diazepam discrimination may be an inappropriate model for studying anxiolytic properties of drugs. Furthermore, the difference between baclofen and diazepam with respect to anxiolytic efficacy is suggested to be related to receptor specificity. The binding of diazepam facilitates GABA_A receptor activity, whereas the binding of baclofen facilitates BABA_B receptors Hence, it is suggested that modulation of GABA_A receptors may be critical in diazepam's blockade of the PTZ stimulus.